1. Field of the Invention
The invention relates to a method of improving the oxidation and corrosion resistance of a superalloy article and to a superalloy article obtained by the process.
The invention is applicable to all kinds of superalloys, and particularly to monocrystalline superalloys and to superalloys having a low grain boundary density and weakly alloyed with hafnium (hafnium concentration below 0.5% by mass).
2. Summary of the Prior Art
The makers of land and aeronautical turbine engines are always faced with demands to increase efficiency and reduce specific consumption. One way of responding to these demands is to increase the temperature of the burnt gases at the turbine inlet. However, this approach is limited by the ability of the turbine parts, such as the distributors or the rotor blades of the high-pressure stages, to withstand high temperatures. Refractory metallic materials called superalloys have been developed for these turbine parts. These superalloys are nickel, cobalt or iron based and provide the component with mechanical strength at high temperatures (creep resistance). At present the burnt gas temperature, which is typically 1600.degree. C. for a modern engine, exceeds the melting point of the superalloys used, and the high-pressure stage blades and distributors are convection cooled by air at 600.degree. C. taken from the compressor stages. Some of the cooling air which flows in the internal channels of the articles is discharged through ventilation apertures in the wall of the article to form a cold air film between the article surface and the hot gases.
In parallel to the adoption of sophisticated cooling techniques several generations of superalloy have been developed with increased creep resistance to meet the need to increase the temperature limit at the turbine inlet. The working temperature limit of these superalloys is of the order of 1050.degree. C.
The improvements in superalloys have been made to the detriment of their oxidation and hot corrosion resistance, which had led to the development of coatings which protect against oxidation and corrosion. There are two kinds of protective coating. The first consists of nickel aluminide (NiAl) type coatings comprising atomic percentage of aluminum between 40% and 55%. These intermetallic coatings may be modified by the addition of chromium and/or a precious metal. The second consists of MCrAlY type metallic coatings where M denotes nickel or cobalt or iron or a combination of these metals. Both these kinds of protective coating form a film of aluminum oxide, called alumina, which insulates the metal below the coating from the external environment.
After the development of superalloys and techniques for cooling rotor blades and distributors, heat barrier coatings constitute the most recent technology for achieving significant temperature gains at the turbine inlet. Heat barrier technology consists of coating superalloy articles with a fine insulating ceramic layer whose thickness can vary from a few tens of microns to several millimetres. In most cases the ceramic layer consists of zirconia stabilised by yttria, which has the advantages of being a poor heat conductor and having good chemical stability at high temperatures. The ceramic layer may be deposited by heat spraying or by electron beam physical vapour deposition, or EB-PVD for short. EB-PVD is the preferred method of making a deposition on the body of blades and distributors, mainly because the coating has a good surface texture and obstruction of the ventilation holes in the articles can be monitored. The ceramic layer deposited by EB-PVD consists of microcolumns perpendicular to the article surface. This microstructure enables the coating to adapt to thermal or mechanical deformations in the plane of the superalloy substrate.
The main difficulty with heat barrier technology is to ensure satisfactory adhesion of the ceramic layer to the article it is required to protect. In contrast to ceramic coatings prepared by hot spraying, adhesion of a ceramic layer deposited by EB-PVD is not mechanical but consists of chemical bonds with the article surface. The ionic conductivity and the porous structure of a zirconia-based ceramic layer is such as to permit, at high temperatures, the diffusion of oxygen from the ambient medium towards the interface with the metallic article, so that the metal oxidises.
If adhesion between the ceramic layer and the superalloy article is to be satisfactory the oxide film formed at the interface between the superalloy and the ceramic layer by EB-PVD must adhere both to the metal of the article and to the ceramic layer, have good mechanical strength, and limit oxidation of the metal below. To increase adhesion of the ceramic layer to the superalloy article it is known to interpose between the superalloy and the EB-PVD ceramic layer a sublayer which serves as a growth site for an alpha alumina film whose thickness varies from a few tenths of a micron to several microns. The EB-PVD heat barrier sublayers used so far are coatings developed to protect superalloys against high-temperature oxidation. These coatings have the property of being alumino-forming, i.e. forming an alumina film in the presence of oxygen at high temperatures. U.S. Pat. Nos. 4,321,311, 4,401,697 and 4,405,659 teach the use of MCrAlY type coatings as a heat barrier sublayer. U.S. Pat. Nos. 4,88,0614, 4,916,022 and 5,015,502 disclose the advantage of using coatings belonging to the aluminide family as a heat barrier sublayer.
It is also known from U.S. Pat. No. 5,427,866 and published European patent application 0718420 deposit the ceramic layer directly on a superalloy base whose surface has been modified by a precious metal of the platinum group. The superalloy surface is modified by deposition of an electrolytic platinum layer several microns thick on the base superalloy, followed by a vacuum diffusion heat treatment at a temperature between 1000.degree. C. and 1150.degree.C. The platinum reacts with the aluminum of the base superalloy to form a complex platinum aluminide incorporating a number of elements including nickel.
It is well known that superalloy oxidation resistance can be improved by the addition of yttrium to the superalloy, the weight percentage of yttrium varying from a few tens of ppm (ppm denoting parts per million) to several percent. Adding yttrium considerably improves the adhesion of the oxide films. Some other elements such as hafnium, zirconium, cerium and in general the lanthanides also help to improve the adhesion of the alumina layers. This effect of adding yttrium and/or related elements, called reactive elements, is exploited in U.S. Pat. No. 5,262,245 which describes a heat barrier coating having a ceramic layer deposited directly on a superalloy covered by an alumina film without the use of a sublayer. The absence of sublayer reduces production costs and weight and gives improved control over the geometry of thin-walled blade bodies.
The beneficial effect on adhesion of the oxide layers achieved by adding yttrium and/or reactive elements is mainly due to the trapping of the sulphur impurity at the core of the alloy in the form of yttrium sulphides or oxysulphides. The sulphur trapped by the addition of reactive elements is not free to move at high temperatures and cannot segregate at the oxide/metal interfaces.
The bad effect of residual sulphur on the adhesion of the alumina layers formed on superalloys has been shown by the experiments of Smialek et al in "Effect of Sulphur Removal on Scale Adhesion to PWA 1480", Metallurgical and Materials Transactions, A Vol. 26A, February 1995. These experiments consisted of submitting to cyclic oxidation MiCrAl specimens which had been desulphurized by heat treatment in hydrogen. The oxidation behaviour of a desulphurized alloy is found to be comparable with that of an alloy doped by the addition of yttrium or other reactive elements. U.S. Pat. No. 5,538,796 describes the deposition of an EB-PVD ceramic layer directly on a base alloy desulphurized to a content of less than 1 ppm and covered by an alumina film without using a sublayer and without adding yttrium to the superalloy. This U.S. Patent specifies that aluminide coatings have a sulphur content which can vary from 8 to 70 ppm, which is a strong argument against using them as EB-PVD heat barrier sublayers on a superalloy whose sulphur content has previously been reduced to less than 1 ppm. However, to improve their creep resistance the new generation superalloys usually include small amounts of aluminum and chromium. These amounts are not enough to ensure the longevity of the alumina layer formed directly on these superalloys without a sublayer, even after the alloy has been given a desulphurizing treatment. The life of the alumina layer in the absence of a sublayer is short because the reservoir of aluminum is low, as is the reactivity of the aluminum in the superalloy. The low chromium content of the superalloy does not enable the chromium to enhance the reactivity of the aluminum.
The various coatings or heat barrier sublayers used to increase adhesion of the ceramic layers deposited on the superalloys and to improve the oxidation resistance thereof are very effective on polycrystalline alloys, but usually perform worse on monocrystalline alloys. Indeed, we have found that the spalling resistance of heat barriers deposited by an EB-PVD process and the oxidation behaviour of the antioxidation coatings is much lower on monocrystalline alloys than on polycrystalline alloys.
By way of example FIG. 1 shows the working life ranges of EB-PVD heat barriers deposited on the polycrystalline superalloys known as IN100 and Hastelloy X and on the monocrystal known as AM1, the superalloys having first been coated with a platinum-modified aluminide sublayer. The alloy AM1 is a nickel based alloy containing, by weight, 7.5% Cr, 6.5% Co, 2% Mo, 8% Ta, 5.5% W, 1.2% Ti and 5.3% Al. The alloy IN100 is a nickel based alloy containing, by weight, 13%-17% Co, 8%-11% Cr, 5%-6% Al, 4.3% to 4.8% Ti, 2% to 4% Mo, 0.7% to 1.2% V, 0.03% to 0.06% Zr, and 0.01% to 0.014% B. The alloy Hastelloy X is a nickel based alloy containing, by weight, 20.5% to 23% Cr, 17% to 20.0% Fe, 8% to 10% Mo, 0.5% to 2.5% Co, and 0.2% to 1.0% W.
The working life of a heat barrier is expressed in terms of the number of heat cycles until spalling of 20% of the surface of the coated specimen occurs. A cycle consists of a step of one hour at 1100.degree. C. with a temperature rise time of 5 minutes and a cooling time to a temperature below 100.degree. C. of 10 minutes.
FIG. 1 shows that the spalling resistance of an EB-PVD heat barrier is less on the monocrystal AM1, whereas the unprotected AM1 has an intrinsic oxidation resistance much greater than that of the polycrystal IN100, which is an alumino-forming superalloy strongly loaded with titanium, and of Hastelloy X which is a chromo-forming alloy. Also, it was observed that protective coatings such as the MCrAlY coatings and the single aluminides modified by chromium or by a precious metal have an oxidation resistance on monocrystal very much less than that observed on polycrystals. Consequently, none of the known coatings used alone or as a heat barrier sublayer has an adequate working life when deposited on a monocrystalline superalloy.
Using scanning electron microscopy it was found that early spalling of an EB-PVD ceramic layer deposited on a monocrystalline superalloy previously coated with a sublayer corresponds to the propagation of a crack at the interface between the alumina film and the metal of the sublayer. This kind of rupture leads to poor adhesion of the oxide film to the metal, which at temperatures above 850.degree. C. may be caused by segregation of the element sulphur at the oxide/sublayer interface.
The sulphur content of the alloy AM1 is between1 and 3 ppm by weight. This content is appreciably lower than that measured in Hastelloy X (20 ppm) and in IN100 (6-10 ppm), yet on these substrates the spalling resistance of the EB-PVD ceramic layer is better.